The proposed research focuses on developing new measures of early child langugae development and parent input from parent-child interaction data, and using those measures to predict children's language and literacy skills in the early elementary years. The purpose of the independent phase (ROO) of this research is to: 1) build on the skills learned during the mentored phase of the award (K99) to develop and validate new measures of child syntax and parent input from transcripts of parent-child interaction, 2) to model growth in these measures during the eariy childhood period looking at relations between parent input and child language development, and 3) to examine the utility of these measures (as well as the vocabulary measures developed during the mentored phase of the award) as predictors of chidlren's later language and literacy skills. The data come from two existing datasets of transcripts from parent-child interaction. The first dataset is from the Language Development Project at the University of Chicago and consists of 90-minute interactions every 4 months between child ages 14-54 months in a sample of 60 families from diverse backgrounds. The second dataset is from the Early Head Start study at Harvard University and consists of 10-minute interactions once a year between child ages 14-36 months in a sample of 108 low-income families. Both samples have additional data using parent report and standardized measures of child language for validating new measures, and both samples have measures of children's literacy skills at kindergarten for use in the final phase of the research. Thus, the overall goals of the entire K99/R00 research program are to: 1) develop measures of children's vocabulary and syntactic skills during eariy childhood that capture rarity of production, 2) better understand the relation between specific parent input measures and child language growth, and 3) further the research on the role of eariy oral language abilities in predicting later literacy skills.